FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010 Burford, et al. Page 1 Evaluation of Friction Stir Weld Process and Properties for Aerospace Application: e-NDE for Friction Stir Processes D. Burford, P. Gimenez Britos, E. Boldsaikhan, and J. Brown National Institute for Aviation Research, Wichita State University, Wichita, KS E-mail: [email protected]Abstract A powerful new friction stir welding e-NDE technique, which is based on process monitoring, shows promise for increasing the accuracy and precision of probability of detection (POD) analyses when compared to conventional inspection techniques. The technique is based primarily on monitoring the F y (transverse) force feedback signal, which has previously been correlated with defect formation. As an e-NDE near real- time inspection technique, force feedback process monitoring adds a second layer of “greenness” to an already extremely “green” process by reducing and potentially elimi- nating the need for secondary inspection operations like X-ray, and ultrasonic inspection steps. In terms of establishing standards and specifications for friction stir technologies, the e-NDE technique featured in this paper will greatly facilitate the establishment of performance based specifications for FSW that will ultimately become the basis of de- veloping design data for FSW joints in multiple structures made from multiple alloys and product forms.
27
Embed
Evaluation of Friction Stir Weld Process and Properties for
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 1
Evaluation of Friction Stir Weld Process and Properties for Aerospace Application:
e-NDE for Friction Stir Processes
D. Burford, P. Gimenez Britos, E. Boldsaikhan, and J. Brown National Institute for Aviation Research,
A powerful new friction stir welding e-NDE technique, which is based on process
monitoring, shows promise for increasing the accuracy and precision of probability of
detection (POD) analyses when compared to conventional inspection techniques. The
technique is based primarily on monitoring the Fy (transverse) force feedback signal,
which has previously been correlated with defect formation. As an e-NDE near real-
time inspection technique, force feedback process monitoring adds a second layer of
“greenness” to an already extremely “green” process by reducing and potentially elimi-
nating the need for secondary inspection operations like X-ray, and ultrasonic inspection
steps. In terms of establishing standards and specifications for friction stir technologies,
the e-NDE technique featured in this paper will greatly facilitate the establishment of
performance based specifications for FSW that will ultimately become the basis of de-
veloping design data for FSW joints in multiple structures made from multiple alloys and
product forms.
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 2
Introduction
Successful implementation of friction stir welding (FSW) is reliant upon basic metal-
working principles applied and controlled at a local level.1 The thermal and mechanical
mechanisms involved in FSW are similar to those found in other metal working proc-
esses such as rolling, extruding, and forging. However, unlike these bulk thermo-
mechanical (TM) processing operations, the highly localized nature of FSW introduces
steep thermal and deformation gradients into the material adjacent to and along the joint
line. Therefore, standards and specifications for friction stir technologies must of ne-
cessity address and account for the localized metal working nature of friction stir
technologies.
e-NDE for FSW: In FSW the side of the weld tool is pressed against the workpiece
in a manner similar to that of machining with the side of an end mill. However, unlike
end mill machining, in FSW the tool design and process parameters are selected such
that the displaced material is captured and reconstituted into the original material – as
opposed to removing it from the work zone in the form of “chips” as is done in machin-
ing. Consequently, there are both similarities as well as dramatic differences in the
dynamic response of the respective tools used in end milling and FSW.
In machining, it is important to clear the cut metal (chips) from the tool at a sufficient
rate to prevent clogging of tool features, namely the flutes, etc. In FSW the opposite is
true. The features of a FSW tool, such as threads, grooves, etc., are expected to be-
come impacted with metal – and thereby maintain a full frontal engagement between
the tool and the material of the workpiece – while in machining only the tool cutting
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 3
edges are expected to be in contact with the workpiece. This full engagement between
the FSW tool and workpiece leads to unique dynamic behavior not typically experienced
in machining.
In machining, advanced control techniques have been investigated for reducing
chatter. For example, Zhang and Sims assessed the ability of “piezoelectric active vi-
bration damping” to arrest chaotic tool behavior.2 To reduce defect formation in FSW
associated with chaotic tool motion, Boldsaikhan,3 and Jene et al.,4 have studied ma-
chine tool-material interactions by monitoring force feedback signals. As these studies
demonstrate, in both machining and FSW, process monitoring may serve as the basis
for reducing chaotic tool behavior and, thereby, provides a means for improving part
quality in both machining and FSW.
In FSW, the tool tends to vibrate or oscillate side-to-side (nominally transverse to
applied loading vector) while under the local dynamic side loading conditions imposed
on the tool at the tool-workpiece interface. In machining, when the tool oscillates in a
chaotic manner, a self-excited vibration phenomenon called “chatter” tends to form,
leaving erratic markings on the newly cut surface. Similar chaotic oscillations in FSW
tend to be associated with the formation of voids within the joint (resulting from the lack
of consistency in the reconsolidation of material along the joint line).
The advancing, rotating FSW tool presses against the material directly ahead of it,
creating a shearing action that extends around the tool front. In a generalized manner,
when the material directly in front of the tool is sufficiently heated under the pressure
and shearing action imposed on it by the advancing FSW tool, thin layers of material are
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 4
transported from the advancing side of the tool to the retreating side of the tool.a This
action is then repeated, with cooler material again being exposed to the leading face of
the rotating, advancing tool.
Figure 1: Schematic cross-section of a generic FSW tool probe located
midway below the tool shoulder and the end of the probe to depict the ide-
alized oscillation of the tool as it advances. Tool rotation is counter-
clockwise and the direction of travel is toward the top of the page. The re-
action forces act on the tool in opposition to the tool motion. A periodic
shearing and movement of metal along the leading edge of the tool – from
the advancing side to the retreating side – results and the tool oscillates
a The advancing side of the tool is the side of the tool where the rotation direction is the same as the travel direction of the tool. The retreating side of the tool is the opposite side where the rotational direc-tion of the tool is opposite the travel direction.
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 5
side-to-side (nominally) in response the primary reaction forces acting on
the leading edge of the FSW tool probe.
This new interface or band of material is again pressed upon until it is sufficiently
heated to be moved along the tool front from the advancing side to the retreating side.
This undulation in metal movement along the leading edge of the tool promotes an os-
cillatory or alternating pattern in both normal and shear forces acting on the tool surface,
which in turn causes the tool to move in a periodic motion, nominally side-to-side, as the
tool is advanced. This process is schematically depicted in Figure 1 (depicting only sim-
plified, idealized frontal force conditions).
Material flow and the associated resultant forces acting upon the tool are actually
much more complex than idealized in the model shown in Figure 1.5 With the tool probe
completely submerged in workpiece, forces act on the probe from all directions in re-
sponse to its dynamic loading environment, the resultant of which may be measured
experimentally.6 The full engagement of the rotating, advancing FSW tool further ag-
gravates its tendency to oscillate in a chaotic manner. Adding to the complexity of FSW
tool oscillatory motion is the spinning motion of the tool shoulder face on the surface of
the workpiece. This tends to cause a walking motion of the end of the tool, which even
further promotes chaotic tool behavior as the tool seeks (or seeks to establish) a center
of rotation on the workpiece surface.
Uniformity in FSW tool oscillations is dependent upon the periodicity (or lack thereof)
in the material flow behavior around the tool front. It is anticipated that the lower the
abruptness in the material heating and shearing cycle, the less likely the process will
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 6
become chaotic in its behavior (action). Selection of tool features and process parame-
ters are expected to contribute to the overall stability of the tool control process.
The ability to monitor the dynamic behavior of FSW tools through force feedback sig-
nals provides an effective way to dramatically reduce or eliminate the inspection costs
associated with secondary inspection techniques such as X-ray, ultrasonic phased array
(UPA), or penetrant inspections. By simply analyzing the force feedback signal of each
weld, this lean and effective e-NDE technique can be utilized to improve production and
quality based directly on recorded weld information. It further offers the potential ability
to actively and adaptively control FSW operations in production. It can also conceivably
be developed to monitor tool wear, optimize design and performance of FSW tools, and
compete different tooling design concepts, etc.
Thermal Components of FSW: The thermal process elements or components of
FSW are typically controlled indirectly (i.e. passively) through the process variables that
most strongly influence them, namely mechanical factors such as spindle speed, travel
speed, and the applied weld tool axial force. Through the influence of these indirect
means, thermal energy is generated during FSW by forcing a rotating, non-consumable
metalworking tool into the joint line between components to be joined. Once stable
processing conditions are established locally, the weld tool is then forcibly translated or
advanced along the joint line to form a consolidated unit.
The energy for conveying material from the advancing to the retreating side of the
weld tool is supplied by the torque and compressive forces of the FSW machine as ap-
plied to the workpiece through the specialized, non-consumable metalworking tool. The
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 7
actual energy imparted to the workpiece by the machine is converted into heat through
mechanical stirring and frictional/shearing interaction between the non-consumable tool
and workpiece. This heat, which is generated in a local but traveling work zone, can be
viewed conceptually as flowing away from the work zone along three generalized heat
sink paths (or conduits):
Path 1: The Spindle Path: including the metalworking (welding) tool, tool holder,
spindle, machine frame, etc.
Path 2: The Workpiece Path: the workpiece, fixture, machine bed, machine
frame, clamps, connecting structure, etc.
Path 3: The Surroundings Path: the atmosphere, applied materials (coolants,
gases, etc.).
Ideally, the distribution of heat flow away from the localized work zone will remain stable
without either a substantial build-up of heat or a substantial loss of energy as the weld
progresses. The level of heat build-up or loss may shift due to, for example, a local
change in the thermal mass of the part and/or fixture (e.g. at a stiffener or with an in-
crease or decrease in section thickness). Or it may result from traveling at a rate faster
than heat can be dissipated along these three paths collectively.
In practice, the proportion of heat that flows along each of these heat sink paths at
any given time can vary widely. Many factors influence the relative heat flux along each
path. For example, in Path 2, the workpiece path, the flux of heat away from the local
work zone is first regulated by the thermal conductivity and heat capacity of the work-
piece and is then regulated by these same properties of the fixturing and supporting
components (e.g. the backing bar). For regularly shaped parts, where the effective
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 8
thermal mass cross-section does not vary over the length of the part, a greater probabil-
ity exists that the process will remain stable throughout the duration of the FSW
process. In contrast, in irregularly shaped parts or setups, which vary in thermal mass
along the direction of the weld (e.g. variations in the joint cross-section), joint properties
can vary substantially as a result of the changing thermal environment (heat sink) in and
around the local work zone if not properly accounted for and addressed.
Edge effects also have the potential for contributing to joint property variation. As
the FSW process progresses toward the end of a workpiece, for example, the heat gen-
erated in the part tends to build up near the end of the part where there is a decreasing
amount of material available to contain the heat generated by the advancing tool. Po-
tential approaches for maintaining a consistent thermal environment as the local work
zone nears the end of a part may include changing process parameters to lessen the
heat input into the joint line in the closeout region of the joint.
Rather than attempting to precisely regulate heat flow during FSW, application de-
velopment work is typically based on a phenomenological approach in which process
parameters are developed for each unique setup and welding system.7 Bounding welds
are usually conducted first to identify a suitable process window limit. Then experimen-
tal design techniques (SPC and DOE) are employed to refine the process window for
optimizing selected joint properties. If changes are made to any of the three general
thermal conduits in the system, the process output (e.g. joint material properties) should
be checked to determine what impact, if any, there may have been as a result of the
change. With such an approach, thermal management may be viewed as more of an
art than a science. However, this approach is often justified where a thorough analysis
FAA Joint Advanced Materials & Structures (JAMS) Center of Excellence 6th Annual Technical Review Meeting, May 19–20, 2010
Burford, et al. Page 9
of the setup is not warranted or deemed tractable given the available program re-
sources.
The actual thermal efficiency of a given FSW process, and the gradients associated
with it, may never be well understood or directly controllable. As such, attempting to es-
tablish repeatable processes through a single rigid process specification (e.g. fixing the
setup, tool, process parameters, weld system, etc.) for all applications is not deemed
necessary or even appropriate. Notwithstanding the complexities involved, perform-
ance specifications, along with the appropriate controlling documents (e.g. welding
performance specifications), provide sufficient control to achieve the ultimate process
goal of fabricating structure that meets engineering requirements.
Mechanical Components of FSW: Unlike the thermal components of the FSW proc-
ess, the mechanical components are typically controlled directly through the FSW
machine capabilities and controls, the selection of the metalworking tool and fixture de-
signs, setting processing speeds and feeds, etc. Because process controls can be set
directly through machine settings and tool designs, defining a process specification
around machine controls may seem to be a straightforward approach to establishing
handbook quality data for FSW. However, the steep gradients introduced by FSW
mean that small variations in input (independent) variables (speed, feed, load, tool de-
sign, etc.) can lead to relatively large variations in local response variables (e.g. thermal